Adaptive copy number evolution in malaria parasites

Copy number polymorphism (CNP) is ubiquitous in eukaryotic genomes, but the degree to which this reflects the action of positive selection is poorly understood. The first gene in the Plasmodium folate biosynthesis pathway, GTP-cyclohydrolase I (gch1), shows extensive CNP. We provide compelling evidence that gch1 CNP is an adaptive consequence of selection by antifolate drugs, which target enzymes downstream in this pathway. (1) We compared gch1 CNP in parasites from Thailand (strong historical antifolate selection) with those from neighboring Laos (weak antifolate selection). Two percent of chromosomes had amplified copy number in Laos, while 72% carried multiple (2–11) copies in Thailand, and differentiation exceeded that observed at 73 synonymous SNPs. (2) We found five amplicon types containing one to greater than six genes and spanning 1 to >11 kb, consistent with parallel evolution and strong selection for this gene amplification. gch1 was the only gene occurring in all amplicons suggesting that this locus is the target of selection. (3) We observed reduced microsatellite variation and increased linkage disequilibrium (LD) in a 900-kb region flanking gch1 in parasites from Thailand, consistent with rapid recent spread of chromosomes carrying multiple copies of gch1. (4) We found that parasites bearing dhfr-164L, which causes high-level resistance to antifolate drugs, carry significantly (p=0.00003) higher copy numbers of gch1 than parasites bearing 164I, indicating functional association between genes located on different chromosomes but linked in the same biochemical pathway. These results demonstrate that CNP at gch1 is adaptive and the associations with dhfr-164L strongly suggest a compensatory function. More generally, these data demonstrate how selection affects multiple enzymes in a single biochemical pathway, and suggest that investigation of structural variation may provide a fast-track to locating genes underlying adaptation.     

A selective sweep driven by pyrimethamine treatment in southeast asian malaria parasites

Malaria parasites (Plasmodium falciparum) provide an excellent system in which to study the genomic effects of strong selection in a recombining eukaryote because the rapid spread of resistance to multiple drugs during the last the past 50 years has been well documented, the full genome sequence and a microsatellite map are now available, and haplotype data can be easily generated. We examined microsatellite variation around the dihydrofolate reductase (dhfr) gene on chromosome 4 of P. falciparum. Point mutations in dhfr are known to be responsible for resistance to the antimalarial drug pyrimethamine, and resistance to this drug has spread rapidly in Southeast (SE) Asia after its introduction in 1970s. We genotyped 33 microsatellite markers distributed across chromosome 4 in 61 parasites from a location on the Thailand/Myanmar border. We observed minimal microsatellite length variation in a 12-kb (0.7-cM) region flanking the dhfr gene and diminished variation for approximately 100 kb (6 cM), indicative of a single origin of resistant alleles. Furthermore, we found the same or similar microsatellite haplotypes flanked resistant dhfr alleles sampled from 11 parasite populations in five SE Asian countries indicating recent invasion of a single lineage of resistant dhfr alleles in locations 2000 km apart. Three features of these data are of especially interest. (1). Pyrimethamine resistance is generally assumed to have evolved multiple times because the genetic basis is simple and resistance can be selected easily in the laboratory. Yet our data clearly indicate a single origin of resistant dhfr alleles sampled over a large region of SE Asia. (2). The wide valley ( approximately 6 cM) of reduced variation around dhfr provides "proof-of-principle" that genome-wide association may be an effective way to locate genes under strong recent selection. (3). The width of the selective valley is consistent with predictions based on independent measures of recombination, mutation, and selection intensity, suggesting that we have reasonable estimates of these parameters. We conclude that scanning the malaria parasite genome for evidence of recent selection may prove an extremely effective way to locate genes underlying recently evolved traits such as drug resistance, as well as providing an opportunity to study the dynamics of selective events that have occurred recently or are currently in progress.     

Amplification of the multidrug resistance gene pfmdr1 in Plasmodium falciparum has arisen as multiple independent events.

The multidrug resistance (MDR) phenotype in mammalian tumor cells can involve amplification of mdr genes that results in overexpression of the protein product termed P-glycoprotein. Chloroquine resistance (CQR) in Plasmodium falciparum has similarities with the MDR phenotype in tumor cells, and some isolates of P. falciparum have amplified levels of the pfmdr1 gene. To investigate the nature and origin of pfmdr1 amplicons, we have cloned large regions of a 110-kb amplicon from the CQR cloned isolate B8 by using the yeast artificial chromosome system. We have identified and sequenced the breakpoints of the amplicon by a novel method employing inverted polymerase chain reaction that is applicable to analysis of any large-scale repeat. We show that the five copies of the amplicon in this isolate are in a head to tail configuration. A string of 30 A's flank the breakpoints on each side of the amplified segment, suggesting a mechanism for the origin of the tandem amplification. Polymerase chain reaction analysis with oligonucleotides that cross the B8 breakpoint has shown in 26 independent CQR isolates, 16 of which contain amplified copies of pfmdr1, that amplification of the pfmdr1 gene in P. falciparum has arisen as multiple independent events. These results suggest that this region of the genome is under strong selective pressure.     

The evolutionary fate of heterogeneous gene duplications: A precarious overdominant equilibrium between environment,sublethality and complementation

Gene duplications occur at a high rate. Although most appear detrimental, some homogeneous duplications (identical gene copies) can be selected for beneficial increase in produced proteins. Heterogeneous duplications, which combine divergent alleles of a single locus, are seldom studied due to the paucity of empirical data. We investigated their role in an ongoing adaptive process at the ace‐1 locus in Culex pipiens mosquitoes. We assessed the worldwide diversity of the ace‐1 alleles (single‐copy, susceptible S and insecticide‐resistant R, and duplicated D that pair one S and one R copy), analysed their phylogeography and measured their fitness to understand their early dynamics using population genetics models. It provides a coherent and comprehensive evolutionary scenario. We show that D alleles are present in most resistant populations and display a higher diversity than R alleles (27 vs. 4). Most appear to result from independent unequal crossing‐overs between local single‐copy alleles, suggesting a recurrent process. Most duplicated alleles have a limited geographic distribution, probably resulting from their homozygous sublethality (HS phenotype). In addition, heterozygotes carrying different HS D alleles showed complementation, indicating different recessive lethal mutations. Due to mosaic insecticide control practices, balancing selection (overdominance) plays a key role in the early dynamics of heterogeneous duplicated alleles; it also favours a high local polymorphism of HS D alleles in natural populations (overdominance reinforced by complementation). Overall, our study shows that the evolutionary fate of heterogeneous duplications (and their long‐term role) depends on finely balanced selective pressures due to the environment and to their genomic structure.     

SYBR Green I and TaqMan quantitative real-time polymerase chain reaction methods for the determination of amplification of Plasmodium falciparum multidrug resistance-1 gene (pfmdr1)

The pfmdr1 gene, which encodes P-glycoprotein homolog 1, has been shown to be a reliable marker of resistance for Plasmodium falciparum related to artesunate and mefloquine combination therapy. The aims of this study are to investigate the copy number of pfmdr1 in P. falciparum isolates collected from the 4 malaria-endemic areas of Thailand (Kanchanaburi, Mae Hongson, Ranong, and Tak) along the Thailand-Myanmar (Burma) border (Thai-Myanmar border) by using SYBR Green I and the standard method TaqMan real-time polymerase chain reaction (RT-PCR) and to compare the efficiency (sensitivity and specificity) of SYBR Green I with TaqMan RT-quantitative (q)PCR methods in determining pfmdr1 gene copy number. Ninety-six blood samples were collected onto filter paper from patients with uncomplicated falciparum malaria who attended malaria clinics in the Kanchanaburi (n = 45), Mae Hongson (n = 18), Ranong (n = 11), and Tak (n = 22) provinces in Thailand. Parasite genomic DNA was extracted from dried blood spots by using QIAcube? automated sample preparation. Pfmdr1 gene copy number was determined by TaqMan (63 samples) and SYBR Green I (96 samples) real-time PCR. Seventy-one (74.0%), 14 (14.6%), 10 (10.4%), and 1 (1%) isolates carried 1, 2, 3, and 4 pfmdr1 gene copies, respectively. Forty-three of 48 (89.6%), 6 of 11 (54.5%), and 3 of 4 (75.0%) samples, respectively, showed agreement with results of 1, 2, and 3 pfmdr1 gene copies as determined by both methods. The efficiency of SYBR Green I in identifying pfmdr1 gene copy number was found to be significantly correlated with that of TaqMan. Considering its simplicity and relatively low cost, SYBR Green I RT-qPCR is therefore a promising alternative technique for the determination of pfmdr1 copy number.     

Decreased prevalence of Plasmodium falciparum resistance markers to amodiaquine despite its wide scale use as ACT partner drug in Zanzibar

ABSTRACT: BACKGROUND: Zanzibar has recently undergone a rapid decline in Plasmodium falciparum transmission following combined malaria control interventions with artemisinin-based combination therapy (ACT) and integrated vector control. Artesunate-amodiaquine (ASAQ) was implemented as first-line treatment for uncomplicated P. falciparum malaria in Zanzibar in 2003. Resistance to amodiaquine has been associated with the single nucleotide polymorphism (SNP) alleles pfcrt 76T, pfmdr1 86Y, 184Y and 1246Y. An accumulation of these SNP alleles in the parasite population over time might threaten ASAQ efficacy. The aim of this study was to assess whether prolonged use of ASAQ as first-line anti-malarial treatment selects for P. falciparum SNPs associated with resistance to the ACT partner drug amodiaquine. METHODS: The individual as well as the combined SNP allele prevalence were compared in pretreatment blood samples from patients with uncomplicated P. falciparum malaria enrolled in clinical trials conducted just prior to the introduction of ASAQ in 2002-2003 (n = 208) and seven years after wide scale use of ASAQ in 2010 (n = 122). RESULTS: There was a statistically significant decrease of pfcrt 76T (96-63%), pfmdr1 86Y (75-52%), 184Y (83-72%), 1246Y (28-16%) and the most common haplotypes pfcrt/pfmdr1 TYYD (46-26%) and TYYY (17-8%), while an increase of pfcrt/pfmdr1 KNFD (0.4-14%) and KNYD (1-12%). CONCLUSIONS: This is the first observation of a decreased prevalence of pfcrt 76T, pfmdr1 86Y, 184Y and 1246Y in an African setting after several years of extensive ASAQ use as first-line treatment for uncomplicated malaria. This may support sustained efficacy of ASAQ on Zanzibar, although it was unexpected considering that all these SNPs have previously been associated with amodiaquine resistance. The underlying factors of these results are unclear. Genetic dilution by imported P. falciparum parasites from mainland Tanzania, a de-selection by artesunate per se and/or an associated fitness cost might represent contributing factors. More detailed studies on temporal trends of molecular markers associated with amodiaquine resistance are required to improve the understanding of this observation.     

Common 5' beta-globin RFLP haplotypes harbour a surprising level of ancestral sequence mosaicism

Blocks of linkage disequilibrium (LD) in the human genome represent segments of ancestral chromosomes. To investigate the relationship between LD and genealogy, we analysed diversity associated with restriction fragment length polymorphism (RFLP) haplotypes of the 5' beta-globin gene complex. Genealogical analyses were based on sequence alleles that spanned a 12.2-kb interval, covering 3.1 kb around the psibeta gene and 6.2 kb of the delta-globin gene and its 5' flanking sequence known as the R/T region. Diversity was sampled from a Kenyan Luo population where recent malarial selection has contributed to substantial LD. A single common sequence allele spanning the 12.2-kb interval exclusively identified the ancestral chromosome bearing the "Bantu" beta(s) (sickle-cell) RFLP haplotype. Other common 5' RFLP haplotypes comprised interspersed segments from multiple ancestral chromosomes. Nucleotide diversity was similar between psibeta and R/T-delta-globin but was non-uniformly distributed within the R/T-delta-globin region. High diversity associated with the 5' R/T identified two ancestral lineages that probably date back more than 2 million years. Within this genealogy, variation has been introduced into the 3' R/T by gene conversion from other ancestral chromosomes. Diversity in delta-globin was found to lead through parts of the main genealogy but to coalesce in a more recent ancestor. The well-known recombination hotspot is clearly restricted to the region 3' of delta-globin. Our analyses show that, whereas one common haplotype in a block of high LD represents a long segment from a single ancestral chromosome, others are mosaics of short segments from multiple ancestors related in genealogies of unsuspected complexity.     

Sequence Variation within the KIV-2 Copy Number Polymorphism of the Human LPA Gene in African,Asian, and European Populations

Amazingly little sequence variation is reported for the kringle IV 2 copy number variation (KIV 2 CNV) in the human LPA gene. Apart from whole genome sequencing projects, this region has only been analyzed in some detail in samples of European populations. We have performed a systematic resequencing study of the exonic and flanking intron regions within the KIV 2 CNV in 90 alleles from Asian, European, and four different African populations. Alleles have been separated according to their CNV length by pulsed field gel electrophoresis prior to unbiased specific PCR amplification of the target regions. These amplicons covered all KIV 2 copies of an individual allele simultaneously. In addition, cloned amplicons from genomic DNA of an African individual were sequenced. Our data suggest that sequence variation in this genomic region may be higher than previously appreciated. Detection probability of variants appeared to depend on the KIV 2 copy number of the analyzed DNA and on the proportion of copies carrying the variant. Asians had a high frequency of so-called KIV 2 type B and type C (together 70% of alleles), which differ by three or two synonymous substitutions respectively from the reference type A. This is most likely explained by the strong bottleneck suggested to have occurred when modern humans migrated to East Asia. A higher frequency of variable sites was detected in the Africans. In particular, two previously unreported splice site variants were found. One was associated with non-detectable Lp(a). The other was observed at high population frequencies (10% to 40%). Like the KIV 2 type B and C variants, this latter variant was also found in a high proportion of KIV 2 repeats in the affected alleles and in alleles differing in copy numbers. Our findings may have implications for the interpretation of SNP analyses in other repetitive loci of the human genome.     

Invasion of gene duplication through masking for maladaptive gene flow

Gene duplication can increase an organism's ability to mask the effect of deleterious alleles present in the population, but this is typically a small effect when the source of the genetic variation is mutation. Migration can introduce orders of magnitude more deleterious alleles per generation and may therefore be an important force acting on the structure of genomes. Using formal analytical methods, we study the invasion of haplotypes containing two copies of the resident allele, assuming that a single-locus equilibrium is already established in a continent-island model of migration. Provided that the immigrant allele can be completely masked by multiple functional gene copies, a new duplication will deterministically spread so long as duplicate haplotypes are, on average, fitter than single-copy haplotypes. When fitness depends on the number of immigrant allele copies and their masking ability then the threshold for invasion depends on the rate of immigration and the rate of recombination between the gene copies. Results from several special cases, including formation of protein dimers and Dobzhansky-Muller incompatibilities, suggest that duplications can invade in a wide range of selection regimes. We hypothesize that duplication in response to gene flow may provide an explanation for the high levels of polymorphism in gene copy number observed in natural populations.     

Using a pericentromeric interspersed repeat to recapitulate the phylogeny and expansion of human centromeric segmental duplications

Despite considerable advances in sequencing of the human genome over the past few years, the organization and evolution of human pericentromeric regions have been difficult to resolve. This is due, in part, to the presence of large, complex blocks of duplicated genomic sequence at the boundary between centromeric satellite and unique euchromatic DNA. Here, we report the identification and characterization of an approximately 49-kb repeat sequence that exists in more than 40 copies within the human genome. This repeat is specific to highly duplicated pericentromeric regions with multiple copies distributed in an interspersed fashion among a subset of human chromosomes. Using this interspersed repeat (termed PIR4) as a marker of pericentromeric DNA, we recovered and sequence-tagged 3 Mb of pericentromeric DNA from a variety of human chromosomes as well as nonhuman primate genomes. A global evolutionary reconstruction of the dispersal of PIR4 sequence and analysis of flanking sequence supports a model in which pericentromeric duplications initiated before the separation of the great ape species (>12 MYA). Further, analyses of this duplication and associated flanking duplications narrow the major burst of pericentromeric duplication activity to a time just before the divergence of the African great ape and human species (5 to 7 MYA). These recent duplication exchange events substantially restructured the pericentromeric regions of hominoid chromosomes and created an architecture where large blocks of sequence are shared among nonhomologous chromosomes. This report provides the first global view of the series of historical events that have reshaped human pericentromeric regions over recent evolutionary time.     

Size matters: non-LTR retrotransposable elements and ectopic recombination in Drosophila

The Drosophila melanogaster genome contains approximately 100 distinct families of transposable elements (TEs). In the euchromatic part of the genome, each family is present in a small number of copies (5-150 copies), with individual copies of TEs often present at very low frequencies in populations. This pattern is likely to reflect a balance between the inflow of TEs by transposition and the removal of TEs by natural selection. The nature of natural selection acting against TEs remains controversial. We provide evidence that selection against chromosome abnormalities caused by ectopic recombination limits the spread of some TEs. We also demonstrate for the first time that some TE families in the Drosophila euchromatin appear to be only marginally affected by purifying selection and contain many copies at high population frequencies. We argue that TEs in these families attain high population frequencies and even reach fixation as a result of low family-wide transposition rates leading to low TE copy numbers and consequently reduced strength of selection acting on individual TE copies. Fixation of TEs in these families should provide an upward pressure on the size of intergenic sequences counterbalancing rapid DNA loss through small deletions. Copy-number-dependent selection on TE families caused by ectopic recombination may also promote diversity among TEs in the Drosophila genome.     

The P-glycoprotein homologues of Plasmodium falciparum: Are they involved in chloroquine resistance?

Chloroquine has been the mainstay of antimalarial chemotherapy but the rapid spread of resistance to this important drug has now compromised its efficacy. The mechanism of chloroquine resistance has not been known but recent evidence from Plasmodium falciparum, the causative agent of the most severe form of human malaria, suggested similarities to the multidrug resistance phenotype (MDR) of mammalian tumour cells which is mediated by a protein molecule termed P-glycoprotein. Two mdr genes (pfmdr1 and pfmdr2) encoding P-glycoprotein homologues have been identified in P. falciparum and one of these (pfmdr1) has several alleles that have been linked to the chloroquine resistance phenotype. In contrast analysis of a genetic cross between chloroquine-resistant and -sensitive P. falciparum has suggested that the genes encoding the known P-glycoprotein homologues are not linked. This review outlines the similarities of the chloroquine resistance phenotype with the MDR phenotype of mammalian tumour cells and explores the possible role of the pfmdr genes.     

Population diversity and antibody selective pressure to Plasmodium falciparum MSP1 block2 locus in an African malaria-endemic setting

Background

Genetic evidence for diversifying selection identified the Merozoite Surface Protein1 block2 (PfMSP1 block2) as a putative target of protective immunity against Plasmodium falciparum. The locus displays three family types and one recombinant type, each with multiple allelic forms differing by single nucleotide polymorphism as well as sequence, copy number and arrangement variation of three amino acid repeats. The family-specific antibody responses observed in endemic settings support immune selection operating at the family level. However, the factors contributing to the large intra-family allelic diversity remain unclear. To address this question, population allelic polymorphism and sequence variant-specific antibody responses were studied in a single Senegalese rural community where malaria transmission is intense and perennial.

Results

Family distribution showed no significant temporal fluctuation over the 10 y period surveyed. Sequencing of 358 PCR fragments identified 126 distinct alleles, including numerous novel alleles in each family and multiple novel alleles of recombinant types. The parasite population consisted in a large number of low frequency alleles, alongside one high-frequency and three intermediate frequency alleles. Population diversity tests supported positive selection at the family level, but showed no significant departure from neutrality when considering intra-family allelic sequence diversity and all families combined. Seroprevalence, analysed using biotinylated peptides displaying numerous sequence variants, was moderate and increased with age. Reactivity profiles were individual-specific, mapped to the family-specific flanking regions and to repeat sequences shared by numerous allelic forms within a family type. Seroreactivity to K1-, Mad20- and R033 families correlated with the relative family genotype distribution within the village. Antibody specificity remained unchanged with cumulated exposure to an increasingly large number of alleles.

Conclusion

The Pfmsp1 block2 locus presents a very large population sequence diversity. The lack of stable acquisition of novel antibody specificities despite exposure to novel allelic forms is reminiscent of clonal imprinting. The locus appears under antibody-mediated diversifying selection in a variable environment that maintains a balance between the various family types without selecting for sequence variant allelic forms. There is no evidence of positive selection for intra-family sequence diversity, consistent with the observed characteristics of the antibody response.     

The rice retrotransposon Tos17 prefers low-copy-number sequences as integration targets

The rice retrotransposon Tos17 is highly activated by tissue culture. To evaluate the impact of transposition of Tos17 on the rice genome and examine its utility for insertional mutagenesis, more than 100 sequences flanking newly transposed Tos17 copies were characterised. The 5-bp target-site duplications flanking Tos17 did not show any consensus sequence, and preferred nucleotides, A/T and G/C, were only found at the second and third nucleotides from both ends of the target site duplications, respectively, indicating that Tos17 has relatively low target-site specificity at the nucleotide sequence level. Integration targets were widely distributed over the chromosomes; however, preferential integration into the sucrose synthase 2 gene and into Tos17 itself was demonstrated by PCR screening using pooled DNA prepared from the mutant population. Hybridisation studies indicated that Tos17 preferentially integrates into low-copy-number regions of the genome. In agreement with this result, about 30% of flanking sequences examined showed significant homology to known genes. Taken together, these results show that Tos17 can have a significant impact on the rice genome and can be used as a tool for efficient insertional mutagenesis.     

Maintenance of the copy number of retrotransposon MDG3 in the Drosophila melanogaster genome]

The genomes of laboratory stocks and natural population of Drosophila melanogaster contain 8-12 copies of retrotransposon MDG3 detected by in situ hybridization. Construction of genotypes with decreased MDG3 copy number using X-chromosome and chromosome 3 free of MDG3 copies results in appearance of hybrid genomes carrying up to 7-10 copies, instead of 2-4 copies expected. New MDG3 copies are detected in different genome regions, including the 42B hot spot of their location. The chromosomes, where new clusters of MDG3 were observed, carry conserved "parental pattern" of MDG1 arrangement. The data obtained suggest the existence of genomic mechanism for maintenance of retrotransposon copy number on a definite level.     

The Complex Genetic Structure of Sugarcane Limits Identification of Additional SNP-Defined Simplex Alleles in Microsatellite Loci

Cultivated sugarcane possesses a large number of chromosomes (typically >80) which can be organised into eight homology groups, each containing approximately 12 homo(eo)logous chromosomes. Currently, microsatellite (SSR) markers are the most easily used markers for marker-assisted selection and other genetic applications. However, only SSR alleles that segregate as simplex and duplex markers can be incorporated into sugarcane maps using populations of ～300 progeny. Consequently only a subset of possible alleles have been mapped for a given SSR locus and are available for subsequent QTL analyses. Three sugarcane SSR loci, mSSCIR8, mSSCIR17 and mSSCIR18, were amplified, cloned and sequenced from the parents of an Australian sugarcane mapping population, IJ76-514 and . The sequences were examined to identify nucleotide sequence polymorphisms in the flanking regions that could provide additional simplex SSR allele markers. Alignment of the sequences revealed SNP, indel and repeat length variation and the pattern of sequence variation suggested multiple alleles for each SSR, including all of the alleles previously scored by fragment length. While the flanking regions of the SSR loci contained numerous SNPs and indels, none defined new simplex alleles within multiplex fragments and no new SSR simplex alleles were mapped. Furthermore, many of the sequence-defined alleles appear to be spurious and may have arisen from PCR-mediated recombination. These results confirm the difficulties associated with characterising allelic diversity in a polyploid species and the complexity of mapping in sugarcane.     

pfmdr1 mutations contribute to quinine resistance and enhance mefloquine and artemisinin sensitivity in Plasmodium falciparum

The emergence and spread of multidrug resistant Plasmodium falciparum has severely limited the therapeutic options for the treatment of malaria. With ever-increasing failure rates associated with chloroquine or sulphadoxine-pyrimethamine treatment, attention has turned to the few alternatives, which include quinine and mefloquine. Here, we have investigated the role of pfmdr1 3' coding region point mutations in antimalarial drug susceptibility by allelic exchange in the GC03 and 3BA6 parasite lines. Results with pfmdr1-recombinant clones indicate a significant role for the N1042D mutation in contributing to resistance to quinine and its diastereomer quinidine. The triple mutations S1034C/N1042D/D1246Y, highly prevalent in South America, were also found to enhance parasite susceptibility to mefloquine, halofantrine and artemisinin. pfmdr1 3' mutations showed minimal effect on P. falciparum resistance to chloroquine or its metabolite mono-desethylchloroquine in these parasite lines, in contrast to previously published results obtained with 7G8 parasites. This study supports the hypothesis that pfmdr1 3' point mutations can significantly affect parasite susceptibility to a wide range of antimalarials in a strain-specific manner that depends on the parasite genetic background.